MCPcopy Create free account
hub / github.com/alibaba/GraphScope / graphlearn_torch

Function graphlearn_torch

python/graphscope/client/session.py:1682–1713  ·  view source on GitHub ↗
(
    graph,
    edges,
    edge_weights=None,
    node_features=None,
    edge_features=None,
    node_labels=None,
    edge_dir="out",
    random_node_split=None,
    num_clients=1,
    manifest_path=None,
    client_folder_path="./",
    return_pyg_remote_backend=False,
)

Source from the content-addressed store, hash-verified

1680
1681
1682def graphlearn_torch(
1683 graph,
1684 edges,
1685 edge_weights=None,
1686 node_features=None,
1687 edge_features=None,
1688 node_labels=None,
1689 edge_dir="out",
1690 random_node_split=None,
1691 num_clients=1,
1692 manifest_path=None,
1693 client_folder_path="./",
1694 return_pyg_remote_backend=False,
1695):
1696 assert graph is not None, "graph cannot be None"
1697 assert (
1698 graph._session is not None
1699 ), "The graph object is invalid" # pylint: disable=protected-access
1700 return graph._session.graphlearn_torch(
1701 graph,
1702 edges,
1703 edge_weights,
1704 node_features,
1705 edge_features,
1706 node_labels,
1707 edge_dir,
1708 random_node_split,
1709 num_clients,
1710 manifest_path,
1711 client_folder_path,
1712 return_pyg_remote_backend,
1713 ) # pylint: disable=protected-access

Callers

nothing calls this directly

Calls 1

graphlearn_torchMethod · 0.80

Tested by

no test coverage detected